TSHR

Tile Operation Diagram

TSHR tile operation

Introduction

Elementwise shift-right of two tiles.

Math Interpretation

For each element (i, j) in the valid region:

\[ \mathrm{dst}_{i,j} = \mathrm{src0}_{i,j} \gg \mathrm{src1}_{i,j} \]

Assembly Syntax

PTO-AS form: see PTO-AS Specification.

Synchronous form:

%dst = tshr %src0, %src1 : !pto.tile<...>

IR Level 1 (SSA)

%dst = pto.tshr %src0, %src1 : (!pto.tile<...>, !pto.tile<...>) -> !pto.tile<...>

IR Level 2 (DPS)

pto.tshr ins(%src0, %src1 : !pto.tile_buf<...>, !pto.tile_buf<...>) outs(%dst : !pto.tile_buf<...>)

C++ Intrinsic

Declared in include/pto/common/pto_instr.hpp:

template <typename TileData, typename... WaitEvents>
PTO_INST RecordEvent TSHR(TileData& dst, TileData& src0, TileData& src1, WaitEvents&... events);

Constraints

  • Intended for integral element types.
  • The op iterates over dst.GetValidRow() / dst.GetValidCol().

Examples

#include <pto/pto-inst.hpp>

using namespace pto;

void example() {
  using TileT = Tile<TileType::Vec, uint32_t, 16, 16>;
  TileT x, sh, out;
  TSHR(out, x, sh);
}

ASM Form Examples

Auto Mode

# Auto mode: compiler/runtime-managed placement and scheduling.
%dst = pto.tshr %src0, %src1 : (!pto.tile<...>, !pto.tile<...>) -> !pto.tile<...>

Manual Mode

# Manual mode: bind resources explicitly before issuing the instruction.
# Optional for tile operands:
# pto.tassign %arg0, @tile(0x1000)
# pto.tassign %arg1, @tile(0x2000)
%dst = pto.tshr %src0, %src1 : (!pto.tile<...>, !pto.tile<...>) -> !pto.tile<...>

PTO Assembly Form

%dst = tshr %src0, %src1 : !pto.tile<...>
# IR Level 2 (DPS)
pto.tshr ins(%src0, %src1 : !pto.tile_buf<...>, !pto.tile_buf<...>) outs(%dst : !pto.tile_buf<...>)